THEJOURNAL
OF BIOLOGICAL
CHEMISTRY
0 1988 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 263,No. 21,Issue of July 25,pp. 10151-10154, 1988
Printed in U.S.A.
The Mechanism of Glycogen Synthetase as Determined by Deuterium
Isotope Effects and PositionalIsotope Exchange Experiments*
(Received for publication, December 21, 1987)
Sung Chun Kim, Amar N. Singh, andFrank M. RaushelS
From the Departments of Chemistry and Biochemistry & Biophysics, Texas A&M University, College Station, Texas 77843-3255
Thereaction mechanism for glycogen synthetase
from rabbitmuscle was examined by a-secondary deuterium isotope effects and positional exchange experiments. Incubation of glycogen synthetasewith [B'802,aj3-'SO]UDP-Glc did not result in any detectable
positional isotope exchange from the 6-nonbridge position to the anomeric oxygen of the glucose moiety.
Glucono-1,5-1actonewas found to be a noncompetitive
inhibitor versus UDP-Glc. The kinetic constants, Ki.
and Kii,were found to be 9 1 f 4 PM and 0.70 f 0.09
mM, respectively. Deoxynojirimycin was a nonlinear
inhibitor at pH 7.5. The a-secondary deuterium
isotope
effects were measured with [1-2H]UDP-Glcby the direct comparison method. The isotope effects on V,,,
and V,,,,JK were found to be 1.23 f 0.04 and 1.09 f
0.06, respectively. The inhibitory effects by gluconolactone and deoxynojirimycon plus the largea-secondary isotope effect on V,,, have been interpreted to
show that an oxocarbonium ion is an intermediate in
this reaction mechanism. The lack of a detectable positional isotope exchange reaction in the absence of
glycogen suggests the formation of a rigid tight ion
pair between UDP and the oxocarbonium ion intermediate.
covalent glycosyl-enzyme (11) adduct as illustratedbelow.
Ho1
The technique of positional isotope exchange (PIX)' has
been applied to a variety of enzyme-catalyzed reactions in the
search for transient intermediates (2-7). The PIX technique
can be used to probe the mechanism of enzymes that direct
the attack of a nucleophile at the anomeric center of UDPGlc. Singh et al. (7) synthesized [p-'802,ap-'80]UDP-Glc(111)
and used this compound to measure the positional
isotope
exchange rate for the interconversion
of a P-nonbridge oxygen
and the anomericoxygen in the reactioncatalyzed by sucrose
synthetase.
"
O
f
I
0
(111)
Glycogen synthetase catalyzes the transfer of a glycosyl
residue from UDP-glucose to the C-4 hydroxylgroup of a
growing glycogen primer with the formation
of an a1,4linkage
as illustratedbelow:
FI
b0--6°C-Urd
I
0
- "
-
O
b
FT FT
C P L P - U r d
I
0
I
0
(IV)
The expectedexchange reaction by sucrose synthetase did not
occur in the absenceof fructose. The additionof 2,5-anhydromannitol, an analog for the active conformation of fructose,
did not stimulate the positional
isotope exchange. This result
UDP-Glc + [glycogen], = UDP + [glycogen],,,
(1) indicates that either the scissile carbon-oxygen bond is not
broken until the fructose is bound to the enzyme or the pRecentkineticstudiesusingequilibriumisotopeexchange
have indicated that glycogen synthetase has a rapid equilib- phosphoryl group of UDP is unable to rotateat a kinetically
rium random Bi Bi mechanism (1). The stereochemical out- significant rate.
a-Deuterium isotope effects and the tight binding by glucome of the overall chemical reaction is one of retention of
configuration at C-1 of the glucose moiety. This observation cono-1,5-lactone and deoxynojirimycin have been used prewould strongly suggest that thecarbon-oxygen bond of UDP- viously as evidence in support of an oxocarbonium ion interof enzymesthat involve the
Glc is completely broken prior to the nucleophilic attack by mediateinthemechanisms
transfer of a glycosyl group t o some acceptor substrate. The
the incomingglycogen. Probable intermediates formed in such
4H3half-chair confora mechanism could either be an oxocarbonium ion (I) or a sp2 character at C-1 and the dominant
mation are thought to be the primary reasons
for the inhibition by gluconolactone of enzymatic reactions involving an
* This work was supported in part by National Institutes of Health oxocarbonium ion intermediate (8, 9). Deoxynojirimycin is a
Grant GM-33894 and Robert A. Welch Foundation Grant A-840. We
acknowledge with thanks financial support by the Board of Regents powerful inhibitor of a-glucosidase (10). Hosie and Sinnott
of Texas A&M University. The costs of publication of this article (9) have proposed a 2,5Bconformation for the oxocarbonium
were defrayed in part by the payment of page charges. This article intermediate in thea-glucosidase reaction. Deoxynojirimycin
must therefore be hereby marked "advertisement" in accordance with canadoptthisconformation
(9). Dahlquist et al. (11, 12)
18 U.S.C. Section 1734 solely to indicate this fact.
$ Recipient of National Institutes of Health Research Career Development Award DK-10366. To whom correspondence should be
' The abbreviations used are: PIX, positional isotope exchange;
addressed: Dept. of Chemistry, TexasA&M University, College Sta- Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic
acid; MES, 4tion, TX 77843.
morpholineethanesulfonicacid.
10151
Mechanism of Glycogen Synthetase
10152
showed that the nonenzymatic acid-catalyzed hydrolysis of 7.5, and 10 mM glucose 6-phosphate in avolume of 12 ml for various
phenyl-@-D-glucopyranoside,which is thought to have a glu- lengths of time a t 25 "C. The reaction was terminated by centrifugacopyranosyl cation as an intermediate in the mechanism, has tion of the reaction mixture through a CF-25 Centriflo ultrafiltration
membrane cone. The filtrate was diluted to 250 ml, and the pH was
an a-deuterium isotope effect ( k H / k D )= 1.13; the nonenzy- adjusted to approximately 8.0. The oxygen-18-labeled UDP-Glc was
matic base-catalyzed hydrolysis, which is considered to react loaded onto a Whatman DE-52 anion exchange column (1.5 X 50 cm)
) 1.03. and was eluted with a 2.0-liter gradient from 10 to 250 mM
via an sN2mechanism, has an isotope effect ( k H / k D =
The lysozyme mechanism, where kH/kg with phenyl-4-0-(2- triethylamine/HCO;, pH 7.5. The fractions containing UDP-Glc were
pooled and evaporated to dryness. The residue was dissolved in a 3.0acetamide-2-deoxy-@-~-glucopyranosyl)-~-D-glucopyranoside
is 1.11, is consistent with an oxocarbonium ion intermediate ml solution containing 100 mM EDTA, 150 mM Tris buffer, pH 9.0,
and 25% D20. Identical experiments were also conducted at pH 7.5
or transition state (12).
induce thePIX
using a-cyclodextrin (0.17 M ) in anattemptto
In this paper, oxygen-18-labeledUDP-Glc has been used to reaction by the binding of other ligands to the active site. The
probe for the existence of such an intermediate in the reaction distribution of oxygen-18 was determined using 31PNMR spectroscatalyzed by glycogen synthetase under a variety of experi- copy as described previously (7).
Inhibition Studies-The incubation medium contained 100 mM
mental conditions. Secondary deuterium isotope effects and
inhibition studies by transition stateanalogs were also utilized Hepes buffer, pH 7.5, 6.0 mM glucose 6-phosphate, 6.0 mM MgCl',
2.0 mg/mlglycogen, and 0.01 units ofglycogen synthetase in a
to elucidate the probable mechanism for this transformation. reaction volume of 0.3 ml. The concentration of UDP-Glc was varied
from 0.3 to 3.0 mM with variable amounts of glucono-1,5-lactone.
The rate of formation of UDP was monitored via high performance
liquid chromatography with a Whatman anion exchange column
Potassium cyanate was obtained from Alpha Chemical Co. [l-*H] using 0.125 M KH2P0, as the
eluting buffer. The datawere fit to the
D-Glucose (V) was purchased from MSD Isotopes Co. K H Z P ' ~was
~,
equation for noncompetitive inhibition.
synthesized from PC& andoxygen-18-labeledwater (97%,Cambridge
Isotope Laboratory) according to themethod of Risley and Van Etten
VA
V=
(2)
(13). NMR analysis of the labeled phosphate indicated >95% incorK(l + I/KJ + A(1 + I/KJ
poration of oxygen-18. Nojirimycin was a gift of Professor C.-H.
Wong (Texas A&M University). Deoxynojirimycin was synthesized
The inhibitory effect by deoxynojirimycin was measured by the
from nojirimycin according to the method of Inouye et al. (14). All same method using a fixed concentration of UDP-Glc (1.0 mM) and
other reagents were acquired from either Sigma or Aldrich. [p-l802,aj3- 100 mM Hepes buffer at pH 7.5 or 100 mM MES buffer, pH 6.0.
'80]UDP-glucose (111) was synthesized according to the method of
Measurement of Deuterium Isotope Effects-Secondary deuterium
Singh et al. (7). [l-'HIUDP-Glc (VIII) was synthesized enzymatically isotope effects of [1-2H]UDP-Glcon glycogen synthetase were determined by the direct comparison of the initial velocities. The initial
from [l-2H]D-glucoseaccording to Scheme I.
velocities were measured by coupling the formation of UDP with the
Synthesis of [l -2H]~-Glucose 6-Phosphate(VI)-Hexokinase was
used to phosphorylate [l-2H]glucose at C-6 with ATP. 10 mM ATP, reactions catalyzed by pyruvate kinase and lactate dehydrogenase.
5.0 mM [1-2H]glucose,5.0 mM MgS04, and30 units of yeast hexoki- Reaction mixtures (3.0 ml) contained 100 mM Hepes buffer, pH 7.5,
nase were incubated in a volume of 100 ml containing 20 mM Tris 6.0 mM glucose 6-phosphate, 6.0 mM MgCl,, 10 mM phosphoenolbuffer, pH 7.5. The formation of [I-*H]glucose 6-phosphate was pyruvate, 2.0 mg/ml glycogen, 0.2mM NADH, 500 units of pyruvate
monitored spectrophotometrically at various times using glucose-6- kinase, 100 units of lactate dehydrogenase, and various concentraphosphate dehydrogenase and NAD+. The reaction was terminated tions of UDP-Glc or [l-'H]UDP-Glc. The reaction was initiated by
after 70 min by centrifuging the reaction mixture through a CF-25 the addition of 0.05 units of glycogen synthetase and monitored by
Centriflo ultrafiltration membrane cone (Amicon Corp.). The filtrate following the change in the NADH concentration at 340 nm. The
was diluted to 500 ml and was applied to a Whatman DE-52 anion isotope effects on V and V/K were obtained according to Equation 3
exchange column (1.5 X 50 cm). The [l-*H]glucose6-phosphate was (151,
eluted using a 2.0-liter gradient from 10 to 200 mM triethVA
7.5. An isolated yield of477 umol was
vlamine/HCO; buffer,. DH
V=
(3)
K ( l FiEv,K) + A(l + F;Ev)
obtained (95%).
Synthesis of [l -'H]Uridine Diphosphoglucose (VZII~-[1-2H]UDPare
Glc was svnthesized by the combined activities of phosphoglucomu- where Fiis the fraction of isotope substitution, and EVand EVIK
tase and UDP-Glc pyrophosphorylase. The reaction mixture con- the isotope effects minus 1 on V and V/K, respectively.
31P
Nuclear Magnetic Resonance Meas~rements-~~P
NMR spectra
tained 2.4 mM [l-2H]glucose 6-phosphate, 4.5 mM UTP, 1.0 mM
were obtained on a Varian XL-400 multinuclear spectrometer operMgSO,, 100 mM Tris buffer, pH 7.5, 0.025 mM glucose 1,6-bisphosating at a frequency of 162 MHz. Typical acquisition parameters were
phate, 500 units of inorganic pyrophosphatase, 500 units of UDP-Glc
6000
Hzsweep width, 15 ps pulse width (45"), and broad-band proton
pyrophosphorylase, and 1000 units of phosphoglucomutase in a voldecoupling. All spectra were referenced to an internal standard of
ume of 200 ml. The formation of UDP-Glc was monitored via high
phosphate at pH 9.0. Up to 10,000 transients were collected and
performance liquid chromatography. The reaction mixture was
Fourier-transformed.
quenched by passing the reaction mixture through an Amicon PM30 ultrafiltration membrane. The filtrate was diluted to 500 ml and
was applied to a Whatman DE-52 anion exchange column (1.5 X 50
RESULTS AND DISCUSSION
cm). The [l-'H]UDP-Glc was eluted with a 2.0-liter gradient from 10
to 300 mM triethylamine/HCOi, pH 7.5. The isolated yield was 350
The timing andthe sequence of the bond-making and bondrmol (73%). The proton NMR spectrum of the [l-'H]UDP-Glc
breaking
steps in the reaction catalyzed by glycogen syntheindicated that theincorporation of deuterium was >95%.
Positional Isotope Exchange Reactions-Glycogen synthetase was tase were examined via a combination of PIX, inhibition, and
incubated with 1.0 mM [/3-'s02,01~'80]UDP-Glc,50 mM Hepes, pH secondary deuterium isotope effect experiments. The PIX
MATERIALS AND METHODS
+
0
HO
(VI)
0
0-P-0
0
-HobD
3
I
0
0
0-P-0-P-Urd
0
0
(VI11
(VIII)
SCHEME
1. Synthesis of [ l-'H]uridine diphosphoglucose. 1, hexokinase, ATP, M e ; 2, phosphoglucomutase;
3, UDPG pyrophosphorylase, UTP, Mg+.
(V)
Mechanism of Glycogen Synthetase
experiments were designed to determine whether the anomeric carbon-oxygen bond of UDP-Glc was broken prior to
the addition of glycogen to the enzyme active site. If the
scissile bond of UDP-Glc can be broken in the absence of
glycogen, then incubation of [~-'802,ap-'80]UDP-Glc with
glycogen synthetase would result in the gradual appearance
of an oxygen-18 atom in the C-0-Pbridging oxygen position.
However, this positional isotope exchange reaction would only
occur if the P-phosphoryl group of the resulting UDP could
rotate freely. The positional isotope exchange reaction can be
detected by the appearance of a new 31PNMR resonance for
the p-P of UDP-Glc that is 0.015 ppm downfield from the
startingmaterial (7). Shown in Fig. lA is the 31P NMR
spectrum of [p-'s02,ap-'80]UDP-Glc (111). The high level of
oxygen-18 incorporation at the indicated positions accounts
for the observation of only one doublet. Shown in Fig. 1B is
the 31PNMR spectrum for the p-P of the labeled UDP-Glc
(111) thatis in isotopic equilibrium with the positionally
exchanged UDP-Glc (IV). The difference inthe observed
chemical shift for the two doublets is due to the larger "0
isotope shift from the P-nonbridging oxygens in I11 than for
the bridging oxygen-18 in IV. Shown in Fig. 1C is the 8-P of
oxygen-18 labeled UDP-Glc (111) that had been incubated
with glycogen synthetase for 100 h. No significant exchange
of label from the 8-nonbridge to theanomeric oxygen position
could be detected. The upper limit for the ratio of the V,,,
for the glycogen synthetase reaction relative to theestimated
maximal PIX rate is greater than 10,000. Identical results
were also obtained when a-cyclodextrin was added to the
reaction mixture. The a-cyclodextrin was used as a dead-end
B
h
n
10153
mimic of glycogenin an attempt toinduce glycogen synthetase
into the proper conformational state necessary for C - 0 bond
cleavage. The a-cyclodextrin contains the a1,4 glucosyl linkage but does not have a free hydroxyl at C-4. Therefore, this
compound cannot serve as a primer for glycogen synthesis.
When 60 mM a-cyclodextrin was included in an assay mixture
(glycogen = 1 mg/ml; UDP-glucose = 1.0 mM) the initial
velocity of glycogen synthesis was reduced by 64%, and thus
a-cyclodextrin is able to bind to glycogen synthetase at high
concentrations. Kokesh and Kakuda (16)have previouslyused
a- or /3-cyclodextrin to induce a PIX reaction in the starch
phosphorylase reaction.
Glucono-1,5-lactone (IX) and deoxynojirimycin (X) were
tested as inhibitors of the glycogen synthetase-catalyzed reaction. These compounds have been previously used as transition-state or intermediate-state analogs of the postulated
oxocarbonium ion.
HO 1
HO 1
(1x1
(X)
Fig. 2 shows that glucono-1,5-lactone is a strong noncompetitive inhibitor uersw UDP-Glc in the reaction catalyzed
by glycogen synthetase. The kinetic constants from a fit of
the data toEquation 2 are Ki. = 0.091 f 0.004 mM and Kii =
0.70 f 0.09 mM. McVerry and Kim (17) have previously
reported the inhibitory effect of glucono-1,5-lactone on the
activity of glycogen synthetase and have suggested that an
intermediate with the glucosyl moiety in a half-chair conformation exists during the reaction sequence. The inhibition by
deoxynojirimycin is more complicated. At pH values higher
than 7.5, the Dixon plot (Fig. 3) for the inhibition by deoxynojirimycin is second order ratherthan first order. This
indicates that there may be more than one site on glycogen
synthetase that interactswith deoxynojirimycin. However, at
pH 6.0 the Dixon plot for the inhibition by deoxynojirimycin
is linear. Sevall and Kim (18)have shown that ATP interacts
with both the UDP-Glc and glucose 6-phosphatesitesin
addition toits regulatory site. Deoxynojirimycin probably
binds at theglycogen site in addition to the UDP-Glc site,
The secondary deuterium isotope effects exhibited by [ l -
I
I
1
l/UDP-Glc
FIG. 1. "P NMR spectra of the 8-P of UDP-glucose. A, [p's0~,~j3-180]UDP-Glc;
B, equilibrium mixture ofI11 and IV; C, [p'800,,~fl-180]UDP-Glc
(111) after incubation with glycogen synthetase
for 100 h.
3
2
(mM"
4
)
FIG. 2. The inhibitory effect of glycogen synthetase by glucono-1.6-lactone. Concentration of gluconolactone: A, 0 mM; B,
0.05 mM; C, 0.10 mM; D, 0.20 mM. Additional details are given in the
text.
Mechanism of Glycogen Synthetase
10154
transition state. D-Glucono-1,5-~actonehas predominantly a
4H3half-chair form (XIV), but Walaszek et al. (19) found by
proton NMR spectroscopy that about 20% of D-glucono-1,5lactone has a 'f5Bconformation in the solution (19).
H:q
/OH
6 - 0 .
Ho OH
2
1
OH
OH
3
Deoxyno J 1 r 1 rnyc i n (rnM)
FIG. 3. Inhibition of glycogen synthetase by deoxynojirimycin. A, pH 7.5; B , pH 6.0.
c
40L
2o
2
4
6
8
10
l/UDP-Glc
12 1 4
(mM"
16 18 20
)
FIG. 4. Double-reciprocal plot for the deuterium isotope effect by [l-"HIUDP-Glc.A, UDP-Glc; B , [1-2H]UDP-Glc.
'HIUDP-Glc were measured in an attempt to determine the
degree of bond hybridization change at the anomeric center
during enzymatic transformation. The secondary deuterium
isotope effects exhibited by [l-'HIUDP-Glc on glycogen synthetase were 1.23 +. 0.04 and 1.09 f 0.06 for Vmaxand Vmax/
I&, respectively. The data areillustrated in Fig. 4. The large
value of 1.23 for the secondary deuterium isotope effect on
Vmaxcan be combined with the inhibitory effect by gluconolactone and deoxynojirimycin to suggest that an oxocarbonium ion intermediate is involved in the reaction mechanism.
Hosie and Sinnott (9) have reported the a-deuterium isotope effects of aryl a-glucopyranosides and a-D-glucopyranosyl pyridium salts with yeast a-glucosidase; isotope effects of
"V = 1.01 and "( V/K) = 0.969 are observed for p-nitrophenyl
a-glucopyranosides,
= 1.22 and "( V/K) = 1.018 for the 4bromoisoquinolinium a-glucopyranosides, and "V = 1.15 and
"( V / K ) = 1.085 for the pyridium a-glucopyranosides salts (9).
The N-glucoside substrates have a different ground state
conformation, a ' S g twist type (XI), than the 0-glucosides
which have a normal 4C1chair form (XII). Hosie and Sinnott
(9) have concluded that a noncovalent event, fast in the case
of the N-glucosides but slow in the case of the 0-glucosides,
precedes the bond-breaking step. Moreover, the inhibition by
deoxynojirimycin, which has a 'z5B conformation (XIII), but
no inhibition by castanospermine which does not, suggests
' conformation in the
that the intermediate adopts the 3' B
4H3 2,5B
(XIII)
(XIV)
The larger a-secondary deuterium isotope effect on V than
on V / K suggests that thereis a high commitment for catalysis
once the substrate has bound to glycogen synthetase (15).
This commitment could arise if UDP-Glc is "sticky" (ie. kOff
is slow relative to Vmax).Alternatively, a fast, but essentially
irreversible, conformational change may precede the kinetically slow bond-breaking step. In conclusion, inhibition of the
glycogen synthetase catalyzed reaction by gluconolactone and
deoxynojirimycin and thelarge a-deuterium isotope effect on
Vmaxby [l-'H]UDP-Glc strongly supports an oxocarbonium
ion intermediate. However, the conformation adopted by this
intermediate cannot be deduced from the available data. The
lack of a detectable positional isotope exchange reaction in
the absence of glycogen suggests the formation of a rigid ion
pair between UDP and the oxocarbonium ion intermediate.
REFERENCES
1. Gold, A. M. (1980) Biochemistry 19,3766-3772
2. Midelfort, C. F., and Rose, I. A. (1976)J. Biol. Chem. 251,58815887
3. Wimmer, M. J., Rose, I. A., Powers, S. G., and Meister, A. (1979)
J. Biol. Chem. 254, 1854-1859
4. Hilscher, L. W., Hanson, C. D., Russell, D. H., and Raushel, F.
M. (1985) Biochemistry 2 4 , 5888-5893
5. Raushel, F. M., and Villafranca, J. J. (1980) Biochemistry 19,
3170-3174
6. Von der Saal, W., Anderson, P. M., and Villafranca, J. J. (1985)
J. Bid. Chem. 260,14993-14997
7. Singh, A. N., Hester, L. S., and Raushel, F. M. (1987) J. Biol.
Chem. 262,2554-2556
8. Legler, G., Sinnott, M. L., and Withers, S. G . (1980) J. Chem.
SOC. Perkin Trans.
II 2,1376-1383
9. Hosie, L., and Sinnott, M. L. (1985) Biochem. J. 226,437-446
10. Legler, G., and Julich, E. (1984) Carbohydr. Res. 128,61-72
11. Dahlquist, F. W., Rand-Meir, T., and Raftery, M. A. (1968) R o c .
Natl. Acad. Sci. U. S. A. 61, 1194-1198
12. Dahlquist, F. W., Rand-Meir, T., and Raftery, M.A. (1969)
Biochemistry 8,4214-4221
13. Riseley, J. M., and Van Etten, R. L. (1978) J. Labelled Compd. &
Radiopharm. 15,533-538
14. Inouye, S., Tsuruka, T., Ito, T., and Niida, T. (1968) Tetrahedron
24,2125-2144
15. Cleland, W. W. (1982) CRC Crit. Rev. Biochern. 13,385-428
16. Kokesh, F. C., and Kakuda, Y. (1977) Biochemistry 16, 24672473
17. McVerry, P. H., and Kim, K.-H. (1974) Biochemistry 13, 35053511
18. Sevall, J. S., and Kim, K.-H. (1971) J. Biol. Chem. 246, 29592964
19. Walaszek, Z., Horton, D., and Ekiel, J. (1982) Carbohydr. Res.
106,193-201